Mitochondria play a central role in the maintenance of the naive state of embryonic stem cells. Many details of the mechanism remain to be fully elucidated. Solute carrier family 25 member 36 (Slc25a36) might regulate mitochondrial function through transporting pyrimidine nucleotides for mtDNA/RNA synthesis. Its physical role in this process remains unknown; however, Slc25a36 was recently found to be highly expressed in naive mouse embryonic stem cells (mESCs). Here, the function of Slc25a36 was characterized as a maintenance factor of mESCs pluripotency. Slc25a36 deficiency (via knockdown) has been demonstrated to result in mitochondrial dysfunction, which induces the differentiation of mESCs. The expression of key pluripotency markers (Pou5f1, Sox2, Nanog, and Utf1) decreased, while that of key TE genes (Cdx2, Gata3, and Hand1) increased. Cdx2-positive cells emerged in Slc25a36-deficient colonies under trophoblast stem cell culture conditions. As a result of Slc25a36 deficiency, mtDNA of knockdown cells declined, leading to impaired mitochondria with swollen morphology, decreased mitochondrial membrane potential, and low numbers. The key transcription regulators of mitochondrial biogenesis also decreased. These results indicate that mitochondrial dysfunction leads to an inability to support the pluripotency maintenance. Moreover, down-regulated glutathione metabolism and up-regulated focal adhesion reinforced and stabilized the process of differentiation by separately enhancing OCT4 degradation and promoting cell spread. This study improves the understanding of the function of Slc25a36, as well as the relationship of mitochondrial function with naive pluripotency maintenance and stem cell fate decision.
Pluripotent stem cells (PSCs) exist in at least two distinct pluripotency states, which have been defined as naive and primed. The naive state has been considered as the ground state of pluripotency and as an ideal tool to generate specific cell types for clinical and basic research, with broader and more robust developmental potential than the primed state . The pluripotency state of stem cells is dynamic. The transition between the naive and primed state can be achieved as a result of the heterogeneous expression of pluripotent genes, such as Nanog , Esrrb [3,4], and Klf4 [4,5]. Tbx3 was found to be one of the key heterogeneously expressed genes in mESCs. Previously obtained data of our group suggests that Tbx3 plays a central role in the naive state maintenance of mESCs. Studying the differential genes due to the heterogeneous expression of Tbx3 in mESCs is not only important to learn more about the naive state regulation in mESCs but also increases the necessary information to acquire naive human embryonic stem cells (hESCs), which are widely considered as an ideal tool for regenerative medicine [6–12].
This study focused on genes related to mitochondria. Mitochondria play a central role in cellular energy production and intermediate metabolism [13–31]. They initiate signaling pathways in response to cellular metabolic and genetic stress and alter nuclear gene expression that causes changes in cell function [32–34]. Many studies indicated that mitochondria are crucial for pluripotency maintenance and stem cell fate decision [35–43]. Mitochondria in PSCs display unique characteristics such as the number of mitochondria, their morphology, activity, and mtDNA integrity. Mitochondria of naive ESCs in both mice and humans were observed by transmission electron microscopy. They were spherical, with poorly developed cristae and electron-lucid matrix. In both primed mESCs and hESCs, mature mitochondria show elongated morphology and more developed cristae compared with their naive state [39,44–49]. In addition, the mitochondrial function and metabolic profile differed remarkably between naive and primed pluripotent states in hESCs [13,15,39,50,51]. An improved understanding of mitochondria in naive state and the primed-to-naive transition will be helpful to find the key mitochondria-related regulators of the naive pluripotency state and to identify mitochondria ‘marker genes’ of the stem cell state.
Solute carrier family 25 member 36 (Slc25a36) was found to be highly expressed in naive mESCs in our study on the heterogeneity of mESCs (data not published). Given the pyrimidine nucleotides transporter activity, Slc25a36 is crucial to maintain normal functions of mitochondria. This study investigated the role of Slc25a36 in the pluripotency maintenance of mESCs.
Slc25a36 (NM_138756) is a member of the nuclear-encoded solute carrier 25 (SLC25) superfamily, also known as the mitochondrial carrier family (MCF). They transport inorganic anions, amino acids, carboxylic acids, nucleotides, metabolites, and cofactors between mitochondria and the cytosol through the membrane. The inner membrane of mitochondria is impermeable for most of the small hydrophilic molecules. Thus, mitochondrial carrier proteins are indispensable for energy generation, amino acid synthesis and degradation, the synthesis of mitochondrial DNA (mtDNA), RNA, and protein, mitochondrial function, and other basic cellular functions [52,53]. SLC25A36 imports/exports pyrimidine nucleotides into and from mitochondria  with preference for cytosine and uracil (deoxy) nucleoside mono-, di-, and triphosphates by uniport and antiport mechanisms. It also transports guanine but not adenine (deoxy) nucleotides [55–57]. Mitochondria contain multiple copies of their own genome, mtDNA, which encodes 13 polypeptides of the mitochondrial respiratory chain that are responsible for mitochondrial oxidative phosphorylation (OXPHOS), as well as two rRNAs and 22 tRNAs for mitochondrial protein synthesis . These transport steps of SLC25A36 are essential for both synthesis and breakdown of mtDNA and RNA, mitochondrial genome maintenance, and mitochondrial function. Although the substrates of Slc25a36 have been identified, its physiological function and the underlying mechanism remain unknown. Only few studies have addressed Slc25a36 so far [59,60]. This study shows that Slc25a36 is important for the maintenance of the undifferentiating state of mESCs. The suppression of Slc25a36 led to mitochondrial dysfunction and consequent differentiation of mESCs. In addition, the glutathione metabolism and focal adhesion pathways participated in this differentiation. The findings offer further insight into the physiological roles of Slc25a36 and integrated mitochondrial function with naive pluripotency maintenance and stem cell fate decisions.
Materials and methods
Mouse E14 ESCs were grown on gelatin-coated plates in mESC complete medium: DMEM (Invitrogen) supplemented with 15% fetal bovine serum (FBS), 2 mM GlutaMax (Invitrogen), 0.1 mM ß-mercaptoethanol (Sigma), 0.1 mM sodium pyruvate, Non-Essential Amino Acids (NEAA, Invitrogen), 1% penicillin/streptomycin, and 1000 IU/ml Leukemia Inhibitory Factor (Millipore). RmESCs were cultured on feeders in mESC complete media, supplemented with 1 µM PD0325901 and 3 µM CHIR99021. Mouse embryonic fibroblast and HEK293FT cells were cultured in DMEM (Invitrogen) with 10% FBS, 2 mM GlutaMax (Invitrogen), and 1% penicillin/streptomycin. The culture of mEpiSCs and mEpiLC were maintained on Fibronectin-coated plates in mEpiSC growth medium as previously described by Brons et al. .
ESC to EpiSC differentiation
According to the naive-to-primed transition protocol described by Guo et al. , E14 mES cells were plated on fibronectin-coated plates at a density of 5.6 × 104 cells/cm2 and cultured in mESC complete medium for 24 h. Then, the medium was changed to the mEpiSC growth medium (N2B27 containing Activin A and Fgf2). Thereafter, cells were maintained in EpiSC culture conditions and passaged every 2–3 days with 5 mg/ml Collagenase II.
shRNA design, plasmid construction, and transfection
For RNA interference, two shRNAs to target 19 base-pair gene-specific regions were designed using the software BLOCK-iT™ RNAi Designer (Life Technology). Oligonucleotides were cloned into pSuper-puro (BglII and HindIII sites; Oligoengine) and cloned into the pLVTHM vector (MluI and Cla I sites, Addgene; 12247), respectively. pSuper-puro and pLVTHM were used to construct shRNA expression against luciferase as negative controls. cDNA of Slc25a36 was amplified and cloned into PiggyBac vectors (MluI and PacI sites) under CAG promoter. All transfection of plasmids into mESCs and 293FT cells were performed using Lipofectamine 2000 (Invitrogen).
Lentiviral production and infection
293FT cells (Invitrogen) were transfected using Lipofectamine 2000 with the following plasmids: pLVTHM, p8.91, and pVSVG. 72 h after transfection, the viral supernatants were collected and filtered through 0.45 µm sterile filters, followed by the concentration of lentivirus particles with PEG8000 Virus Precipitation Solution (5×) overnight at 4°C. For infection, cells were plated at a density of 1 × 105 cells per well of a six-well plate and were infected with lentivirus suspension containing 2 µg/ml polybrene for 48 h.
Alkaline phosphatase staining
The alkaline phosphatase (AP) staining was performed using an Alkaline Phosphatase Detection Kit (Millipore, Temecula, CA, U.S.A.) according to the manufacturer's instruction.
Fluorescence-activated cell sorting
Cells transfected with pLVTHM-shRNAs were collected after trypsinization; then, washed with PBS followed by resuspension of the cells in the fresh mESC growth medium. Before fluorescence-activated cell sorting (FACS) was performed on a FACS Calibur (BD Biosciences), the cells were filtered. The sorted cells were replated and cultured on the plate in mESC complete medium.
RNA isolation and qRT-PCR
The total RNA of each sample was extracted using the RNeasy® Mini Kit (Qiagen). cDNA was synthesized using MMLV Reverse Transcriptase (Promega), and qRT-PCR was performed with the Light Cycler® 480 Instrument (Roche), using 2× RealStar Power SYBR Mixture (Genstar). The specificity of primer was confirmed via agarose gel electrophoresis and melting-curve analysis. A housekeeping gene (either β-actin or GAPDH) was used as an internal standard. The primers used in this study were described in the Supplementary Table S5.
Western blot analysis
The whole cell lysate was prepared with cell lysis buffer for Western blot and IP containing 1× protease inhibitor cocktail and phosphatase inhibitor cocktail (P0013, Beyotime) followed by centrifugation for 5 min at 12 000 rpm. Thirty micrograms of the whole cell lysate was transferred onto a PVDF membrane and hybridized with SLC25A36 antibody [N1C3] (1:1000, GTX119934, Genetex), Oct-3/4 Antibody (C-10) (1:1000, sc-5279, Santa Cruz), Actin antibody (1:1000, AA128, Beyotime), and Flag antibody (1:1000, AF519, Beyotime). The second antibody used in this study was the following: HRP Goat anti-rabbit IgG (H + L) (A0208, Beyotime) and HRP Goat anti-mouse IgG (H + L) (A0216, Beyotime).
Cells were fixed for 20 min in 4% paraformaldehyde followed by permeabilization with 0.1% Triton X-100 for 15 min at room temperature. Cells were blocked with 5% BSA in PBS and shaken for 1 h at 80 rpm. Then, cells were stained with primary antibodies, followed by the appropriate secondary antibodies conjugated with Alexa Fluor 488 or 594 (Invitrogen). Oct-3/4 Antibody (C-10) (sc-5279, Santa Cruz) was diluted at 1 : 400 and AF519 antibody (Beyotime) was used at 1 : 1000 dilution. Anti-Cdx2 (BioGenex, CDX2-88) was used at 1 : 200 dilution. Images were captured with a fluorescence microscope.
sgRNAs design and knockout clones
The sgRNAs targeting Slc25a36 were designed using CRISPR DESIGN (http://crispr.mit.edu/n). The sgRNAs were located at exon 2 and exon 3. Each pair of oligonucleotides was annealed in annealing buffer and ligated with Bas1-linearized pGL3-U6-sgRNA-puro (from Dr. Xingxu Huang's lab). pGL3-U6-sgRNA-puro and pST1374-Cas9-Bst (from Dr. Xingxu Huang's laboratory) were co-transfected into E14 mESCs with 1 µg/ml puromycin and 10 µg/ml Blasticidin selection. PCR products were amplified using primers covering Slc25a36 targeting sites and were sequenced to identify Slc25a36 gene disruption. After selection with Puromycin and Blasticidin for 3 days, 1000 cells were replated on a 10 cm dish to pick single cells.
Mitochondrial DNA copy measurement
The relative mtDNA amount (mitochondrial DNA amount relative to nuclear DNA) measurement was determined via qPCR using primers for Nd2 (NADH dehydrogenase subunit 2; mitochondrial genome) as previously described . The relative Nd2 copy number was calculated based on the threshold cycle (Ct) as 2−Δ(ΔCt), where ΔCt = CtNd2 − Ctgenome, and Δ (ΔCt) = ΔCTsample − ΔCTcontrol.
Mitochondrial membrane potential determination by FACS
Cells were stained with tetramethylrhodamine methyl ester (TMRM) (20 nM) by directly adding it to mESC complete media and incubated at 37°C for 30 min. After fixing with 4% PFA and washing, the cells were observed by fluorescence microscopy and analyzed by flow cytometry (BD Biosciences) according to the manufacturers’ instructions. After living to stain with TMRM, the fluorescence intensity was observed using a fluorescence microscope and analyzed using Summit 2.0 Software following FACS.
Δψm analysis by operetta high-content imaging system
Cells were cultured on gelatin-coated 96 Black, Clear bottom, TC-treated View Plate (6005182, PerkinElmer) in mESC complete medium, containing Mitotracker Deep Red (100 nM) at 37°C for 30 min. After fixing with 4% PFA and washing, the cells were observed by an Operetta High-Content Imaging System. The images were analyzed with Harmony to determine the mean fluorescence density of the Mitotracker Deep Red per cell.
Determination of intracellular GSH and GSSG/GSH
A total of 10 000 cells were collected by centrifugation at 500 g for 10 min and washed twice with cold PBS; then, resuspended in protein removal reagent and vigorously vortexed. The samples were frozen and thawed twice rapidly using liquid nitrogen and a water bath at 37°C; then, at 4°C for 5 min. The supernatant was collected by centrifugation at 10 000 g for 10 min and used for detection. The total GSH and GSSG were determined using GSH and GSSG Assay Kits (S0053, Beyotime), respectively, according to the manufacturers’ instructions. Results are expressed in nmol GSH/mg protein.
SLC25A36 was expressed fusion with 3XFlag in mESCs using PB-CAG-Flag-Slc25a36-puro vector. The cells were cultured and passaged every 3 days under 1 µg/ml puromycin selection. Then, the whole cell lysates were collected to perform immunoprecipitation using FLAG immunoprecipitation kit (Sigma–Aldrich) according to the manufacturer's instruction. The control was transfected with PB-CAG-Flag-puro vectors.
All data were analyzed in triplicate and expressed as means ± SEM. Differences of P < 0.05 were considered statistically significant. All statistical analyses were performed with SPSS statistic 21 software (U.S.A.).
Slc25a36 highly expressed in naive mESCs
According to the analysis of the SLC25A family with the ChIP-seq and mRNA-seq data of Tbx3 (Supplementary Tables S2 and S3) [64,65], the expression of Slc25a36 decreased in Tbx3-knockdown mESCs (Figure 1A). We verified the expression of Slc25a36 in Tbx3 suppression and overexpression mESCs by qPCR. The results showed that the expression of Slc25a36 increased in Tbx3-overexpressed mESCs (Figure 1B), while it significantly decreased in shTbx3 cells (Figure 1C). This suggests that the expression of Slc25a36 might differ between naive and primed mESCs. To validate this, primed mEpiSC-like cells (mEpiLC) were obtained from E14 ESCs (Supplementary Figure S1); then, the expression in mESCs, mEpiLCs, mouse Epiblast stem cells (mEpiSCs), and mouse trophoblast stem cells (mTSC) was analyzed by qPCR. The results showed that the expression of Slc25a36 was elevated in naive mESC (E14, G4), while it was low in primed stem cells (mEpiLCs and mEpiSCs) (Figure 1D). This suggests that Slc25a36 might function in the pluripotency maintenance of mESCs.
Expression and localization of mouse Slc25a36.
Before further characterization, the expression levels in tissues and intracellular localizations in E14 and red mitochondria ESCs (rmESC) were investigated. Slc25a36 showed high expression in lung, oviduct, and brain (Figure 1E). Two cell types were separately transfected with the PB-CAG-Flag-mSLC25A36 and PB-CAG-Flag control plasmids. Cells were fixed and stained with FLAG antibody. The result indicated non-nuclear localization in E14 cells (Figure 1F). Images of rmESCs expressing mitochondria-targeted DsRed (su9–RFP) indicated the mitochondria localization of the SLC25A36 protein (Figure 1G).
In summary, Slc25a36 were mitochondria localized and highly expressed in naive state mESCs. This suggests that Slc25a36 might play a role in pluripotency maintenance.
Slc25a36 was required to maintain pluripotency of mESCs
RNAi knockdown of Slc25a36 was performed in E14 feeder-free mESCs with both pSuper-Slc25a36 shRNAs and pSuper-mock shRNAs. More than 50% knockdown of genes was achieved with this system (Figure 2A), and SLC25A36 protein expression was also significantly decreased (Figure 2B) after 4 days of selection with puromycin. The differentiation morphology of mESCs was observed in ESC growth conditions. As shown in Figure 2C, the colonies became flat, loose, spreading, and lost smooth edges. After AP staining, control mESC colonies stained bright, sharp, and uniform, while the Slc25a36 knockdown colonies appeared dim, fuzzy, and weak. Similar morphologic changes occurred in Slc25a36 knockdown rmESCs (feeder-dependent) (Supplementary Figure S2). Further analysis of germ layer markers by qPCR showed that the expression of pluripotency genes Pou5f1, Sox2, Nanog, and Utf1 decreased in Slc25a36 knockdown mESCs, while trophoblast-associated markers Cdx2, Hand1, and Gata3 were up-regulated (Figure 2D). Both western blot analysis of cell extracts (Figure 2D) and immunostaining of fixed cells (Figure 2E) using OCT4 Antibody showed decreased expression of OCT4 protein in Slc25a36 knockdown mESCs. In addition, a significant decline in endoderm marker expression (Gata6 and Sox17) was observed. However, no consistent changes of marker genes associated with the ectoderm and mesoderm lineages were observed. When cultured under TSC growth conditions, the differentiation phenotype of Slc25a36 knockdown cells was more obvious (Supplementary Figure S3). Three days post-transfection, the cells were separately fixed for immunostaining of OCT4 and CDX2. The fluorescence strength of OCT4-positive cells decreased (Supplementary Figure S4) and CDX2-positive cells were observed at the edge of differentiated colonies. Nevertheless, none of the CDX2-positive cells were identified in control shRNA-treated ESCs (Figure 3A).
Knockdown of Slc25a36 led to the differentiation of mESCs.
CDX2-positive cells emerged in Slc25a36-deficient mESCs in TSC growth medium.
The differentiation phenotypes in Slc25a36 stable knockdown and knockout mESCs were confirmed. Using lentivirus, stable Slc25a36 knockdown cell lines were obtained termed LV-1 (shRNA1 transfected), LV-2 (shRNA1 and shRNA2 co-transfected), LV-NC (non-targeting shRNA as negative control). We also got Slc25a36 knockout cell lines termed Slc25a36+/− (Slc25a36 heterozygous knockout) and Slc25a36−/− (Slc25a36 homozygous knockout) by the CRISPR/Cas9 technology (Figure 3E) (Sequencing results showed in Supplementary Figure S5). qPCR analysis demonstrated that the expression of Slc25a36 mRNA was inhibited by ∼50% in these cell lines. Pou5f1 expression significantly decreased, while Cdx2, Gata3, and Hand1 increased (Figure 3B). As shown in Figure 3C, the typical colony morphology for undifferentiated mESCs was maintained in control and wild-type mESCs, while Slc25a36-deficient cells showed a diffused and spreading appearance. Many Slc25a36-deficient ESCs differentiated into a flattened trophoblast cell-like morphology. Consistent with transient knockdown cells, small amounts of CDX2-positive cells at the edge of differentiated colonies were observed after cultured in TSC growth medium (Figure 3D). These CDX2-positive cells also showed reduced levels of OCT4 (Figure 3F).
In summary, the roles of Slc25a36 in maintaining the undifferentiated state of mESCs and affecting the stem cell fate decision were confirmed.
Slc25a36 was necessary for mitochondrial function in mESCs
Slc25a36 functioned as pyrimidine nucleotides transporter, which was central for synthesis and breakdown of mitochondrial DNA and RNA, as well as for mitochondrial genome maintenance. Therefore, mitochondria were investigated in Slc25a36-deficient mESCs. As shown in Figure 4A, the relative amount of mtDNA decreased. This indicated that the mitochondrial genome might decrease. The key genes related to mitochondria biogenesis that is central for mt genome maintenance (Tfam, Tfb, Nrfs, and PGC1-α/β) were analyzed by qPCR. Significantly decreased expressions of key regulators (PGC1-α, PGC1-β, Nrf1, Nrf2, Tfa, Tfb2, and PPARα) were observed (Figure 4B). The mitochondrial biogenesis was accompanied by variations in mitochondrial size, number, and mass. The mtDNA content was reported as an indicator of mitochondrial dysfunction . Then, the mitochondrial superstructure was observed in Slc25a36 knockout cells using transmission electron microscopy. Not only the number of mitochondria decreased, but mitochondria also appeared swollen and rounder (Figure 4C), which differed from wild-type mESCs. This implies mitochondrial damage.
Slc25a36 deficiency led to mitochondrial dysfunction.
Consequently, the mitochondrial membrane potential (MMP/Δψm) was assessed using TMRM, which reflects the functional status of mitochondria. Both the weak fluorescence intensity of TMRM under the fluorescence microscope (Supplementary Figure S6) and the left-moving wave obtained by FACs analysis (Supplementary Figure S7) indicated a low Δψm in Slc25a36 knockdown mESCs. In addition, this result was confirmed with MitoTracker Deep Red both in Slc25a36 knockdown (Figure 4D) and knockout cells using Operetta High-Content Imaging System. The mean mitochondrial fluorescence intensity per cell was significantly lower than in control (Figure 4E), which was consistent with the results of knockout cells (Supplementary Figures S8 and S9). These results indicate mitochondrial dysfunction in Slc25a36-deficient mESCs and support the observed impaired mitochondria morphology. In addition, knockdown of Slc25a33, another pyrimidine transporter, led to mESC differentiation (Supplementary Figure S10), with quicker and more obvious differentiation under TSC growth conditions (Supplementary Figure S11). This indicates the importance of mtDNA to pluripotency maintenance.
This suggests that Slc25a36 suppression led to mtDNA depletion, thus resulting in mitochondrial dysfunction, finally causing differentiation.
Profiling and bioinformatics analyses revealed pathway regulating differentiation
To gain additional insight into the global effect and functional roles of Slc25a36 in mESCs, deep sequencing was performed to profile the mRNA expression in Slc25a36 knockdown and control mESCs (Supplementary Table S3). KEGG analysis obtained the signaling pathways involved in differentiation regulating. Differential genes were significantly enriched in Focal adhesion, PI3K–Akt signaling pathway, ECM-receptor interaction, glutathione metabolism, metabolic pathways, and other signaling pathways. The genes of the metabolism including glutathione metabolism, metabolic pathways, and glycerolipid metabolism accounted for the largest proportion. This was followed by genes of disease-related signaling pathways, such as cancer, small cell lung cancer, Huntington's disease (Figure 5A). Furthermore, metabolic pathways were significantly down-regulated in Slc25a36 knockdown cells (Figure 5B) while they were up-regulated in Slc25a36-overexpressing mESCs. The HIF-1 signaling pathway was also up-regulated when overexpressed (Supplementary Figure S12). This indicated that Slc25a36 was closely related to metabolic and mitochondrial function.
GO and KEGG pathway analysis of differential genes.
GO analysis by biological processes demonstrated that up-regulated genes were enriched in those that regulated cell shape, cell differentiation, cell adhesion, and regulation of transcription (DNA-templated) (Figure 5C). Down-regulated genes were enriched in ATP hydrolysis coupled proton transport and glutathione metabolic processes (Figure 5D,E).
Moreover, genes of metabolic pathways were significantly down-regulated (Supplementary Figure S13) including the components of mitochondrial complex I: Ndufc2, Ndufs7, Ndufs8, Ndufa6, and Ndufb9; mitochondrial complex III: Cox6a2, Cox5b, and Cox8a (Figure 6A). qPCR verified the decreased expression of these genes (Figure 6B). In combination with mtDNA depletion, this suggests the declined expression of mitochondrial complexes. Inefficient operation of the mitochondrial electron transport chain (ETC) could account for the lower membrane potential of EpiSC compared with ESC . This result supports the low Δψm and mitochondria dysfunction in Slc25a36-deficient mESCs. This result provides further evidence to explain the phenotype of mitochondrial dysfunction and differentiation in mESCs upon Slc25a36 suppression. Further studies are required to explore the functional relationship between Slc25a36 and these pathways.
Down-regulated glutathione metabolism involved in OCT4 expression decreased.
Down-regulated glutathione enhanced OCT4 protein degradation
Glutathione metabolism pathway showed extremely significant down-regulation both in GO and KEGG enrichment (Figure 6C). qPCR analysis verified the decreased expression of genes in this pathway, including Ggt1, Ggt7, Gstm3, Gstk1, Gpx6, Gpx4, Gstt2, Gstt1, and Idh2 (Figure 6D). These genes encode enzymes that participate in the glutathione metabolism. Glutathione is an antioxidant and exists in both reduced (glutathione, GSH) and oxidized (glutathione disulfide, GSSG) states. The GSH/GSSG ratio reflects the oxidation state. Recent studies reported that GSH was necessary for the regulation of pluripotency and decreased GSH directly led to a rapid down-regulation of OCT4 . Therefore, the levels of GSH and GSSG in Slc25a36 knockdown and control mESCs were tested using both GSH and GSSG Assay Kits. The results showed that GSH levels decreased significantly in Slc25a36 knockdown mESCs, while GSSG levels increased (Figure 6E), indicating a low GSH/GSSG ratio (Figure 6F). We further verified this in E14 mESCs under GluMax (an alternative of l-glutamine) withdrawal. Glutamine was a precursor of glutamate, the building block of GSH . In the absence of GluMax, E14 mESCs differentiated. qPCR analysis showed decreased mRNA expressions of Oct4 and Sox2, while increased expression of Cdx2 (Figure 6G). Weak AP staining and diffused colonies were observed and the protein level of OCT4 decreased (Figure 6H).
In conclusion, the reduction in GSH levels in Slc25a36 knockdown mESCs contributed to the changes in OCT4 protein levels. Lowering OCT4 levels enhanced differentiation and increased endothelial cell sprouting.
Up-regulated focal adhesion promoted cell spread
The focal adhesion pathway was found to be strongly up-regulated in Slc25a36 repressing mESCs (Figure 7A). The expression levels of key genes, such as Itga1, Itga9, Itgb5, Tnc, Cav1, and Thbs1 were verified by qPCR. The result showed that these genes increased (Figure 7B). Recent reports indicated that low focal adhesion signaling promoted the ground state pluripotency of mESCs [70,71]. Strong adhesion of mESCs culture led to cell spread and decreased pluripotency marker genes [72,73]. This suggests that the up-regulation of FAs might regulate cell spreading in Slc25a36 knockdown mESCs. To verify this, PF-573,228, an ATP analog inhibiting focal adhesion kinase (FAK) kinase activity was used. FAK was the major kinase in the FA signaling pathway. It was activated and localized at FAs upon cell adhesion to the extracellular matrix [74,75]. This inhibitor was highly specific for the FAK catalytic activity and inhibited FA reassembly. According to previous reports, treatment of cells with 1 µM inhibited FAK phosphorylation up to 80–90% but did not inhibit cell growth. At a concentration of 10 µM, cell growth was significantly inhibited . In this study, cells were treated with different concentrations of PF-573,228 (1, 3, and 10 µM) for 24–48 h. The obtained result indicates that pharmacological FAK inhibition rescued the differentiation phenotype of Slc25a36 knockdown mESCs. The colonies of Slc25a36 knockdown mESCs partially recovered smooth edges with the treatment of 1 µM of FAK inhibitor for 24 h (Figure 7C). Cell growth inhibition was observed with 3 and 10 µM of PF-573,228. These data suggested that the focal adhesion pathway participated in the regulation of differentiation.
Focal adhesion pathway contribution to cell spreading promotion.
Moreover, immunoprecipitation was performed using the FLAG antibody after overexpression of SLC25A36 fusion with FLAG in mESCs to search interactive proteins (Supplementary Figures S14 and S15). Interestingly, intersection network and functional network analysis showed that predicted interactive proteins were mainly enriched in focal adhesion, cell–cell adhesion, and ribosome functional clusters (Supplementary Table S4; Figure 7D).
SLC25A36 functions as a pyrimidine nucleotide carrier, the physiological function of which still remains unknown. In this study, the role of Slc25a36 was characterized in the pluripotency regulation of mESCs, which had not been reported before.
Slc25a36 suppression was demonstrated to cause a decline of mtDNA, MMP, mitochondria numbers, mitochondria biogenesis, and a swelling morphology, suggesting impairment of mitochondria and mitochondrial dysfunction. All of these effects were associated with the pluripotency of stem cells [37,42,49,77], leading to mESC differentiation. According to a previously published report, the measured half-life of mtDNA in mammalian cells was 10–30 days . MtDNA depletion occurred rapidly without the ongoing process of mtDNA replication and the replication that supports the nucleotide metabolism, eventually causing the failure of mitochondria and even causing cell function. Furthermore, it has been reported that imbalanced mitochondrial dNTP pools not only caused mtDNA depletion and dysfunctional mitochondria but also cellular dNTP depletion, which eventually led to genome instability . Research of mitochondria DNA polymerase gamma catalytic subunit (POLG) knockdown in mESCs corroborated these findings: POLG regulated the mtDNA copy number by DNA methylation . POLG knockdown resulted in reduced OCT4 expression and slightly increased levels of the mesodermal marker Brachyuryb . A report by Kelly et al. demonstrated that altered pluripotent gene expression correlated with the mtDNA copy number in reprogrammed cells . This study showed that the maintenance of proper content of the mitochondrial genome, mitochondrial biogenesis, and mitochondrial network integrity are crucial for the pluripotency maintenance, in addition to the glycolytic metabolic state of PSCs [42,82,83].
Furthermore, the possibility that Slc25a36 regulated other pathways was found. The glutathione metabolism and focal adhesion were involved in the differentiation regulation (Figure 7E).
An imbalance of the glutathione metabolism emerged with a significant decrease in GSH and an increase in GSSG. A decrease in the ratio of GSH to GSSG has been considered as an indication of oxidative stress [32,84]. In the reduced state, the thiol group of cysteine is capable to donate the reducing equivalent (H ++ e−) to a protein to maintain reduced forms. Recent studies reported that after the withdrawal of glutamine, endogenous glutathione (GSH) was depleted. Then, OCT4 cysteine residues required for its DNA binding was oxidized and enhanced OCT4 degradation [68,85]. Here, OCT4 was reported to serve as a metabolic-redox sensor. In addition, GSH was distributed in mitochondria with equal cytosol concentration. This caused a decrease in mitochondrial GSH in Slc25a36-deficient mESCs. Mitochondrial GSH played a critical role in the defense against respiration-induced reactive oxygen species, thus negatively regulating mitochondrial function [86–88]. In summary, the regulation of the metabolic pathway to pluripotent genes during differentiation reinforced the differentiation and formed a metabolism–differentiation–mitochondria dysfunction feedback loop.
Focal adhesions (also cell-matrix adhesions or FAs) were integrin-containing, multi-protein structures that form mechanical links between intracellular actin bundles and the extracellular matrix or substrate in many cell types. Regulatory signals can be transmitted between the extracellular matrix and interacting cells through these links. Genes associated with the focal adhesion pathway were dramatically up-regulated. GO analysis demonstrated that genes in the biological processes of regulating cell shape, cell differentiation, cell adhesion, and regulation of transcription (DNA-templated) were significantly up-regulated. Moreover, the intersection network and functional network analysis of SLC25A36 after immunoprecipitation showed that predicted interactive proteins were mainly enriched in focal adhesion, cell–cell adhesion, and ribosome functional clusters. These data implied that Slc25a36 might regulate focal adhesion or transcription through protein–protein interaction. It remains to be elucidated. Furthermore, whether the regulation of pathways in cell adhesion and transcription was the result of differentiation or Slc25a36 suppression remains to be explored as well. In addition, the flat colonies recovered when the PI3K inhibitor LY294002 was added (data not shown). PI3K was positioned downstream of FAK and reported to regulate ESCs proliferation and self-renewal . KEGG analysis showed significant up-regulation of PI3K–Akt signaling pathway next to FA (Supplementary Figure S16). This suggests that FAK–PI3K–AKT signaling might promote differentiation caused by Slc25a36 suppression. Further research is needed to verify this.
Moreover, the expression of Slc25a36 was heterogeneous in mESC, which is consistent with Tbx3. Metabolisms were up-regulated in Tbx3 overexpressing mESCs, which suggests mitochondrial heterogeneity mESCs. These findings may provide more information on mitochondrial heterogeneity and mitochondrial remodeling during the transition and acquisition of pluripotency.
The current investigation of Slc25a36 also has significant implications for cancer, as well as regenerative and clinical medicine. Many genetic diseases, characterized by mtDNA depletion or increased mtDNA mutation rates, are caused by mutations on those nuclear genes encoding enzymes of nucleotide metabolism [90,91]. Future studies are necessary to learn more about the physiological functions of mitochondrial transporters in Slc25a36 knockout mice. GO and KEGG analysis in this study indicate a potential connection between Slc25a36 and several diseases, such as small cell lung cancer, amoebiasis, and Huntington's disease. Elevated expression of Slc25a36 was found in lung and brain (Figure 1E) and several cancer cell lines. Furthermore, there might be a potential interaction between SLC25A36 and cell adhesion. Cellular interactions with extracellular matrix play vital roles in tumor initiation, progression, and metastasis . It suggests that Slc25a36 might play a role in cancer and other diseases.
During the establishment of pluripotency, mitochondria showed significant changes, such as metabolic shift, mitochondrial function, structure, and maturity to meet specific cellular needs in a pluripotent state. An in-depth understanding of the mechanisms underlying these events and these changes not only provides important insights into mitochondria remodeling and mitochondrial heterogeneity in different pluripotency state but also helps to obtain the naive human stem cell, and specifically understand the pathogenesis of diseases.
clustered regularly interspaced short palindromic repeats
electron transport chain
epiblast stem cell-like cell
epiblast stem cell
fluorescence-activated cell sorting
focal adhesion kinase
fetal bovine serum
human embryonic stem cells
mouse embryonic fibroblast
mouse embryonic stem cells
mitochondrial carrier family
mitochondria membrane potential
pyrimidine nucleotide carrier
pluripotent stem cells
quantitative polymerase chain reaction
revolutions per minute
single guide RNA
tetramethylrhoda mine methyl ester
trophoblast stem cell
J.H. conceived and supervised the project. Y.X. designed and performed the main experiments; Y.W. performed the Tbx3 heterogeneity experiments; L.Z. performed the RNA-seq analysis; B.S. helped with the CRISPR/Cas9 experiments; H.L. performed FAC-sorting experiments; Y.X. and J.H. wrote the paper.
This work was supported by China National Basic Research Program [2016YFA0100202], National Natural Science Foundation of China [31571497 and 31772601], Plan 111 [B12008], and Research Programs from the State Key Laboratories for Agrobiotechnology, China Agricultural University [2017SKLAB1-2 and 2018SKLAB6-20].
The Authors declare that there are no competing interests associated with the manuscript.